Leading Whitespaces of Language Models' Subword Vocabulary Poses a Confound for Calculating Word Probabilities.
Byung-Doh OhWilliam SchulerPublished in: CoRR (2024)
Keyphrases
- language model
- out of vocabulary
- n gram
- spoken document retrieval
- language modeling
- probabilistic model
- retrieval model
- document retrieval
- language modelling
- statistical language models
- spoken term detection
- information retrieval
- query terms
- query expansion
- speech recognition
- ad hoc information retrieval
- test collection
- translation model
- smoothing methods
- context sensitive
- retrieval effectiveness
- word error rate
- cross language information retrieval
- word segmentation
- term dependencies
- relevance model
- pseudo relevance feedback
- term weighting
- text categorization
- vector space model
- bag of words
- web search